Context: 

A recent study reported that Deep learning approaches are important for drug discovery.

Background of Neural Network

• In 1944, artificial neural networks were first proposed by Warren McCullough and Walter Pitts.
• Later, it led to the birth of deep-learning and artificially intelligent systems like ChatGPT.
• Today, deep learning, a branch of artificial intelligence, is being used to discover new antibiotics.
• The second wave of deep learning revolutions started in 1990 with advances in algorithms and computational power revitalising neural networks.

What is a Neural Network?

• A neural network, also known as an Artificial Neural Network (ANN) is a powerful tool in the field of artificial intelligence (AI), inspired by the structure and function of the human brain.
• Building block: They are fundamental units, consisting of interconnected nodes (artificial neurons) arranged in layers.
• Function: Each node processes information and transmits it to others, mimicking the firing of biological neurons.
• Applications: Diverse talks like image recognition, natural language processing, and speech recognition.

Relations Between Deeper Learning and Antibiotics

• Discovering New Drugs: Deep learning can design new antibiotics from scratch, exploring Chemical Models/ Motifs/ Compounds not easily accessible through traditional methods.
• Repurposing Existing Drugs: Enhancing understanding of specific antibiotic research by providing a broader model for enhancing drug development strategies.
  • Deep learning can identify existing drugs with potential antibacterial activity, accelerating their development into new antibiotics.
• Understanding Resistance Mechanisms: Deep learning can help analyse and predict how bacteria develop resistance to existing antibiotics, guiding the development of new ones that are less susceptible.

How Neural Networks Found Antibiotics?

• Researchers at MIT and Harvard used a type of neural network called a Graph Neural Network (GNN) to analyse a massive database of molecules.
  • The GNN highlighted specific substructures within the molecule that contributed to their antibiotic activity.
• These substructures provided valuable insights into the drug discovery process.
• This explainable artificial intelligence approach allows researchers to not only discover new antibiotics but also understand their core process of work, potentially accelerating the development and optimization of these drugs.

News Source: The Hindu and MIT News